As the 21st century approaches and as advances in the treatment of heart disease and stroke mean many are now surviving longer, the number of persons in older age groups is increasingly rapidly. A number of these persons are likely to require care because of illness and frailty, highlighting the need to understand variations in care patterns. Most such care is given in the community, some by unpaid - usually family - helpers, which is defined as informal care, and some by paid assistants, defined as formal care. Reasons for the mix of formal and informal services have been little studied. This is particularly critical with respect to differences by race, which have been largely ignored. Yet variations in assets, obligations, and attitudes that are implicated in family decisions to provide formal and informal care for their aged relatives undoubtedly vary by race. Data to answer research questions on these issues will be collected from a random sample of 700 northeast Ohio caregivers, evenly divided between black and white, selected by random digit dialing, and then personally interviewed in a prospective four year longitudinal design. A subset of the caregivers will be asked to keep a month long diary of caregiving experiences, to enrich the findings with qualitative insights. Retrospective data will also be secured from relatives of patients in a predominantly black and a predominantly white nursing home to provide information on the different pathways to virtually complete formal care. This three-pronged approach provides for a triangulation model of data collection that will be useful in cross-- validating the results. An innovative model dealing with caregivers' decisions about service use will guide the study.